An EUV underlayer directly improves patterning performance by tuning the optical environment beneath the photoresist, improving adhesion, and supporting pattern transfer fidelity at sub-20 nm feature sizes. Without an optimised underlayer, even the most capable EUV photoresist will produce unacceptable levels of line-edge roughness, pattern collapse, and critical dimension variation. The questions below unpack each dimension of underlayer function, from optical design to stochastic noise management and hard mask co-optimisation.
How does an EUV underlayer affect resist performance?
An EUV underlayer affects resist performance by controlling the standing wave pattern within the resist film, improving adhesion at the resist-substrate interface, and providing a chemically compatible surface for post-exposure development. Together, these functions determine how faithfully the aerial image formed by the EUV scanner is translated into a developed resist pattern. A poorly matched underlayer degrades contrast, increases line-edge roughness, and raises the probability of feature collapse during development.
At the optical level, the underlayer modifies the effective reflectance reaching the resist from below. Because EUV light at 13.5 nm interacts with matter differently than deep-UV wavelengths, even small changes in underlayer refractive index or extinction coefficient shift the standing wave intensity distribution inside the resist. This directly affects the exposure dose absorbed at different depths within the film, which in turn influences the acid generation profile in chemically amplified resists or the crosslinking density in metal oxide systems.
At the interface level, the underlayer provides a surface energy and chemical environment that governs how the resist wets and adheres to the substrate. Poor adhesion is a leading cause of pattern delamination during development rinse steps, particularly as feature aspect ratios increase with each technology node. The underlayer also acts as a diffusion barrier, preventing substrate-generated contaminants from quenching the photoacid or interfering with the resist chemistry above.
What optical properties does an EUV underlayer need?
An EUV underlayer needs a carefully tuned complex refractive index, specifically its real part (n) and its extinction coefficient (k), to minimise back-reflected EUV light reaching the photoresist from below. The target optical properties depend on the resist film thickness, the underlying substrate stack, and the specific scanner illumination conditions. In practice, the underlayer functions as an anti-reflective coating tuned to the 13.5 nm wavelength.
At EUV wavelengths, nearly all materials absorb strongly, so the design challenge is not simply eliminating reflection but balancing absorption against the need for a thin, uniform film that does not itself become a source of standing wave variation. Silicon-based underlayers, including silicon-containing anti-reflective coatings (SiARC), are well suited to this because their optical properties at EUV can be adjusted through composition and film thickness. The silicon content also provides etch selectivity advantages during pattern transfer, which is a secondary but important benefit.
Film thickness uniformity is equally critical. Thickness variation across a 300 mm wafer translates directly into reflectance variation, which means the effective dose seen by the resist changes from die to die. This makes deposition process control and the intrinsic coating uniformity of a spin-on underlayer central to its optical performance in production. PiBond’s silicon underlayers are engineered from proprietary monomers and resins to deliver the compositional consistency required for uniform optical performance at wafer scale.
How does an underlayer reduce EUV pattern collapse?
An EUV underlayer reduces pattern collapse by improving resist adhesion to the substrate surface and by enabling thinner resist films that lower the aspect ratio of developed features. Pattern collapse occurs when capillary forces during the development rinse exceed the mechanical restoring force of a resist line, causing it to deflect and contact an adjacent feature. A higher-adhesion underlayer surface raises the threshold at which collapse occurs by anchoring the base of the resist line more firmly.
Thinner resist films are central to the collapse mitigation strategy at advanced nodes. As critical dimensions shrink below 20 nm, the aspect ratio of a resist line becomes the dominant collapse risk factor. By providing sufficient optical contrast and etch selectivity, a well-designed underlayer allows process engineers to reduce resist thickness while maintaining the contrast needed for pattern definition. This directly lowers the aspect ratio and reduces the mechanical vulnerability of each feature during wet processing.
Surface energy tuning is the other lever the underlayer provides. A surface energy matched to the resist formulation promotes uniform wetting during spin coating and reduces the formation of interfacial voids or delamination sites. Some underlayer formulations incorporate adhesion-promoting functional groups that form covalent or strong non-covalent bonds with the resist polymer, providing a measurable improvement in the force required to initiate collapse.
What’s the difference between organic and inorganic EUV underlayers?
Organic EUV underlayers are carbon-based polymer films that offer good spin-coating uniformity and are compatible with a wide range of resist chemistries, but they provide limited etch selectivity relative to the resist above them. Inorganic underlayers, typically silicon-containing or metal oxide materials, offer significantly higher etch selectivity and stronger EUV absorption tunability, making them better suited to aggressive pattern transfer schemes at advanced nodes.
Organic underlayers
Organic underlayers are generally formulated from carbon-rich polymers and are processed at relatively low bake temperatures. Their main advantages are ease of integration, good planarisation over topography, and compatibility with standard resist track equipment. Their limitation is etch selectivity: because they are chemically similar to organic photoresists, achieving high selectivity in the etch step that transfers the pattern into the hard mask or substrate requires careful process optimisation. At nodes where resist budgets are very thin, this selectivity gap becomes a significant constraint.
Inorganic and silicon-containing underlayers
Silicon-containing underlayers, including SiARC materials, introduce silicon into the film matrix, which provides a fundamentally different etch response to oxygen-based plasma chemistries compared with the carbon-rich resist above. This selectivity difference is what allows the underlayer to act as a pattern-transfer intermediate, preserving critical dimension fidelity as the pattern is etched down through the stack. Inorganic metal oxide underlayers extend this principle further, offering even higher etch resistance and the ability to tune optical properties across a wider range. The trade-off is process complexity and the need for precise compositional control during synthesis, which is why in-house monomer and resin manufacturing is a meaningful capability differentiator for underlayer suppliers.
How does the underlayer interact with EUV stochastic effects?
The underlayer interacts with EUV stochastic effects primarily through its influence on the effective photon dose absorbed within the resist and through the quality of the chemical environment it provides at the resist-substrate interface. EUV stochastic effects arise because the number of photons, photoelectrons, and generated acid molecules within any small volume of resist is subject to statistical variation. The underlayer cannot eliminate this fundamental physics, but it can reduce the sensitivity of the patterned feature to that variation.
Back-reflected EUV light from the underlayer contributes to the total dose absorbed in the lower portion of the resist film. If the underlayer reflectance is poorly controlled, dose variation at the bottom of the resist adds to the stochastic noise already present from photon shot noise at the top. An anti-reflective underlayer with stable, uniform optical properties reduces this additional source of dose variation, effectively tightening the dose distribution across the resist volume and improving the statistical consistency of feature formation.
At the interface, outgassing or chemical contamination from the underlayer can locally quench photoacid or alter the development rate in ways that are spatially random. This produces a pattern-dependent roughness contribution that is distinct from photon shot noise but appears as an increase in line-edge roughness in metrology data. Underlayer formulations with low outgassing and well-controlled surface chemistry therefore contribute to stochastic noise reduction in a way that is often underappreciated relative to the optical contribution.
When should the underlayer be co-optimised with the hard mask stack?
The underlayer should be co-optimised with the hard mask stack from the earliest stage of process integration whenever the target critical dimension requires a resist film thin enough that etch selectivity becomes a limiting factor. In practice, this threshold is reached at most advanced logic and memory nodes below 20 nm half-pitch, where the resist budget is insufficient to survive the full etch depth without the underlayer and hard mask acting as a coordinated pattern-transfer system.
Co-optimisation means aligning the underlayer’s etch response, optical properties, and surface chemistry with the specific hard mask material and etch chemistry used downstream. If the underlayer is selected independently of the hard mask, mismatches in etch selectivity or adhesion at the underlayer-hard mask interface can cause pattern distortion, footing, or delamination during the transfer etch. These failures are difficult to diagnose and correct after the individual layers have been qualified separately, which is why integrated stack development is the more efficient path.
The optical design of the underlayer also interacts with the hard mask in reflectance terms. A hard mask with non-trivial EUV reflectance placed beneath the underlayer shifts the effective optical boundary condition the underlayer must compensate for. This means the underlayer thickness and composition optimised against a bare silicon substrate may not be optimal when the same underlayer sits above a spin-on hard mask. Stack-level optical simulation, followed by experimental validation, is the standard approach to resolving this coupling. PiBond’s collaboration with the Paul Scherrer Institute supports exactly this kind of advanced EUV materials and process development across the full patterning stack.
For process integration engineers working on next-generation nodes, selecting underlayer and hard mask materials from a supplier with in-house synthesis and formulation capability offers a practical advantage: stack-level co-optimisation can be addressed at the materials level, not only through process adjustments at the fab. Contact PiBond to discuss your material requirements and how the underlayer and hard mask stack can be developed together for your specific patterning integration.
Related Articles
- How siloxane monomers define performance limits in patterning underlayers
- Semiconductor dielectric materials supplier selection criteria for fab procurement teams
- What drives spin-on hard mask selection for extreme aspect ratio etching
This content was generated with the help of AI and it may contain mistakes